INTERNATIONAL CONFERENCE ON COMMUNICATION, COMPUTER AND POWER (ICCCP'09) MUSCAT, FEBRUARY 15-18, 2009 © SQU-2009 ISSN: 1813-419X Optimum Allocation of Wind Based DG in Unbalanced Rural Network Y. M. Attwa 1 , E.F. El-Saadany 1 , and M. H. Albadi 1 Abstract— Recent development in small renewable/clean generation technologies, such as wind turbines, photovoltaic, fuel cells, and microturbines, has drawn distribution utilities’ attention to the possible change in the distribution system infrastructure and policy by deploying Distributed Generation (DG). In this paper an Optimal Power Flow (OPF) algorithm is used to determine the optimum allocation of wind-based DG in order to minimizing the total power losses. The OPF is formulated as a Mixed Integer Non-Linear Programming (MINLP) problem, with an objective function to minimize the system power losses. The constraints include voltage limits at different buses (slack, generation, and load buses) of the system, feeder capacity, and maximum penetration limit of DG units. Moreover, the capacity factor of wind based DG is calculated using the Rayleigh model of wind speed, represented by the mean wind speed at the potential locations, and wind turbines data. The OPF model is applied to a typical unbalanced system with different scenarios including restrictions on the size and location of DG. The results show that there are different optimum allocations of the DG units based on the constraints of each scenario penetration level. In addition, the results demonstrate that increasing penetration level of DG over a certain threshold can cause negative impacts on the system power losses. Key words—Distributed generation, optimal power flow, unbalanced system, wind energy. I. INTRODUCTION Under the new deregulation policy and with the expected proliferation of DG technologies, a reasonable share of electricity is expected to be supplied by different DG systems. Further, introducing more DG units into the distribution system will not only help in meeting the electricity demand, relief transmission congestion, but will also introduce variety of other benefits. A study by the Electric Power Research Institute (EPRI) indicates that by 2010, 25% of the new generation will be distributed; while, a study by the Natural Gas Foundation concluded that this figure could be as high as 30% [1]. The European Renewable Energy Study (TERES), commissioned by the European Union (EU) to examine the feasibility of EU CO2-reduction goals and the EU renewable energy targets, found that around 60% of the renewable 1 The authors are with the Department of Electrical and Computer Engineering at the University of Waterloo, Ontario, Canada (e-mail: ymoustaf@uwaterloo.ca). energy potential that can be utilized until 2010 can be categorized as decentralized power sources [2]. Optimal operation and planning analysis is often performed via optimization studies. Traditionally, these optimizations are performed based on minimizing the operational cost while maintaining an appropriate security margin. With the introduction of distributed generation in distribution systems, some aspects of the system optimization related to both planning and operation will arise. Some of these aspects are the penetration level and the optimum location of DG that will reduce electrical losses in the distribution system, which, in turns, will reduce the operation cost. Several articles that address the use of artificial intelligence algorithms to optimize DG placement, based on minimizing power losses, have been published [1]–[6]. Reference [1] solves the problem by an exhaustive algorithm, [2] employs the Tabu search method, [3] uses a fuzzy genetic algorithm and analytical approaches are presented in [4]. Other papers considered the cost of power interruptions [5] and minimizing peaks [6]. The objectives of [7] include reduction in T&D losses and improvement of voltage profile of the system, with due consideration of fixed and variable costs. In [8] an optimization model for minimization of losses through constrained power flows and optimal sitting of DG units in a multi-bus distribution network is presented. The main objective is to minimize the line losses subject to meeting generation and transmission constraints and consumer demand. Most of the previous work done for allocating DG in the distribution system assumed that the system is perfectly balanced and the DG sources are fully guaranteed; two assumptions that are totally invalid for practical distribution systems as well as renewable DG. This paper will introduce the DG optimum size and location optimization problem taking into consideration two vital issues: the impact of system unbalance and the uncertainty of non conventional types of DG. In this paper a novel OPF based technique is used to determine the optimum allocation of wind based DG in the distribution system. The problem is formulated as a MINLP, taking into consideration the effect of system unbalance, and the uncertainty of the wind based DG. The technique is applied on a typical unbalanced rural distribution system, with different scenarios. Moreover, a sensitivity analysis is conducted in order to determine the impact of increasing the DG penetration level on system losses. The paper is organized as follows: the next section broadly discusses the main objective of this work and the proposed procedure to accomplish it. Calculation of the capacity factor of wind based DG is explained in section III. Further, section VI presents a practical distribution system case study with the 325